Contents
function [footage, fData] = getFootage(file, varargin)
% getFootage takes a file and returns the footage % Created by Daniel Darby, adapted from extract_image_sequence_data.m by % Isabel Llorente Garcia. % Reads image sequence of tybe .sif, .dv, .tif, .mat data or .mv2d and % returns image data in a useful form (a structure) % % Inputs: % file: file, including .type and full path input file contains the footage % on a sutiable file format % % varargin{1} = desired data series (optional, defults to one) % % % Outputs: % footage: complete set of images from file in a 3D array % fData, structure containing fields: % 'numFrames': number of frames in footage for series. % 'Ysize': vertical (first dimension of matrix) size of image in pixels % for series. % 'Xsize': horizontal (second dimension of matrix) size of image in % pixels for series. % 'ColData': Collection Data, including camera/decetor model & make, % pixel length etc % % % Usage: % [f1, dF1] = getFootage('N:\FRAP\Data\FRAP_data\filename.mvd2'); % this will automatically open the first series of data, to open series n: % [fn, dFn] = getFootage('N:\FRAP\Data\FRAP_data\filename.mvd2', n); % where n is desired series number %
Initialising and Error control:
Sets up initial variaibles and errors.
% Initial stuff necessary for labeling [~, ~, ext] = fileparts(file); %takes care of ID'ing the file type % For setting the series number if nargin < 2 seriesNum = 1; % No series number input provided assume first series. elseif nargin == 2 && isnumeric(varargin{1}) seriesNum = varargin{1}; else error('gF:inarg', 'Too many input arguments or input 2 (desired series number) is not numeric'); end % Contains supplimentary data fData = struct('numFrames', {}, 'Ysize', {}, 'Xsize', {}, 'file', {},'ext', {}, 'seriesNum', {}, 'numSeries', {}); % fData(1).file = file; fData(1).ext = ext; fData(1).seriesNum = seriesNum; numSeries = 1;
For .mvd2 files
requires bfopen.m and loci_tools.jar
if strcmpi('.mvd2', ext) data = bfopen(file, seriesNum); % data is a cell array of cell arrays. numSeries = size(data,1); series = data{seriesNum,1}; thisnumFrames = size(series,1); % vectorised code footage = double(cat(3,series{:,1})); % Add other useful info to final output: fData(1).numFrames = thisnumFrames; fData(1).Ysize = size(footage,1); fData(1).Xsize = size(footage,2); fData(1).camera = data{1,2}; end
VolocityReader initializing N:\FRAP\Data\FRAP_data_MarkMarsh\220611.mvd2
VolocityReader initializing N:\FRAP\Data\FRAP_data_MarkMarsh\220611.mvd2
Reading series #1
............
For .dv files
This needs bfopen.m and loci_tools.jar (see folder bfmatlab):
if strcmpi('.dv', ext) data = bfopen(file); % data is a cell array of cell arrays. % Assumming there is only one series, i.e., that numSeries = size(data,1)=1, we do: series1 = data{1,1}; % this cell array contains an array where each row (first index, p) corresponds to a frame. % First column (series1{p,1}) is the matrix data for the image frame. Second column (series1{p,2}) is the % frame label with path and info. series1_numFrames = size(series1,1); % number of frames in the image sequence. % Each image frame matrix data is series1{p,1}, with real intensity numbers % (not between 0 and 1). % To produce image data in final output form: % Vectorised code footage = mat2gray(double(cat(3,series{:,1}))); % % Loop through frames: % for p = 1:series1_numFrames % frame = series1{p,1}; % frame = double(frame); % to class double. % frame = mat2gray(frame); % footage(p).image = frame; % % image data needs to be of class double for future operations (for findSpotCentre1frame.m). % end % Add other useful info to final output: fData(1).numFrames = series1_numFrames; fData(1).Ysize = size(footage,1); fData(1).Xsize = size(footage,2); fData(1).camera = data{1,2}; end
For .sif files:
if strcmpi('.sif', ext); image_path = sifImagePath0.name; % First get size of .sif image file: (see IO_Input folder, SifFunctions.txt, or page 95 my notebook 1). % See Isabel/myMatlabFiles/IO_input folder. [~, fData(1).numFrames, ImageSize, ~]=GetAndorSifSize(image_path,0); % numFrames is the length of the sequence, the number of frames. % ImageSize is the size of the image, e.g. 512*512 = 262144. % TotalAcquisitionSize is numFrames*ImageSize % read .sif image data. sifData is an array of 1x1 structures, as many % columns as frames on image sequence. Read frames 1 to numFrames: [sifData] = read_sif_data_direct(image_path,fData(1).numFrames,ImageSize,1,fData(1).numFrames); % sifData is a cell array. footage = mat2gray(double(cat(3,sifData{:}.sliceData))); % for p = 1: fData(1).numFrames % frame = sifData{p}.sliceData; % extract frame data which is stored in the field 'sliceData'. % frame = double(frame); % to class double. % frame = mat2gray(frame); % footage(p).image = imrotate(frame,90); % original intensity values (needs to be rotated). % % image data needs to be of class double for future operations (for % % findSpotCentre1frame.m). % end % % Get frame dimensions: fData(1).Ysize = size(footage,1); fData(1).Xsize = size(footage,2); end
For .tif files
if strcmpi('.tif', ext) image_path = tifImagePath0.name; tif_info = imfinfo(image_path); % Add other useful info to final output: numFrames = length(tif_info); % number of frames in sequence. Ysize = tif_info(1).Height; Xsize = tif_info(1).Width; % Make the image be a square image: Ysize = min(Ysize,Xsize); Xsize = min(Ysize,Xsize); % If the frame size is an odd number: if mod(Ysize,2)~=0 % modulus after division Ysize = Ysize - 1; Xsize = Xsize - 1; end % To produce image data in final output form: %vectorised footage = mat2gray(double(cat(3,imread(image_path,1:numFrames)))); % % Loop through frames: % for p = 1: numFrames % % frame = imread(image_path,p); % read frame number p, read full image. % frame = imread(image_path,p,'PixelRegion', {[1,Ysize], [1,Xsize]}); % read only a part of image, to have a square image at the end. % frame = single(frame); % to class single. % frame = double(frame); % to class double. % frame = mat2gray(frame); % footage(p).image = frame; % % image data needs to be of class double for future operations (for findSpotCentre1frame.m). % end fData(1).numFrames = numFrames; fData(1).Xsize = Xsize; fData(1).Ysize = Ysize; end
For .mat files
if strcmpi('.mat', ext) image_path = matImagePath0.name; load(image_path); % load mat file. % The mat file should contain a matrix of size frameSizeY x frameSizeX % x numFrames. file_info = whos('-file',image_path); % this returns a structure with fields, eg.: % name: 'A_image_sequence' % size: [400 400 100] % bytes: 128000000 % class: 'double' % global: 0 % sparse: 0 % complex: 0 % nesting: [1x1 struct] % persistent: 0 % Get useful info for final output: numFrames = file_info.size(3); % number of frames in sequence. Ysize = file_info.size(1); Xsize = file_info.size(2); % Make the image be a square image: Ysize = min(Ysize,Xsize); Xsize = min(Ysize,Xsize); % If the frame size is an odd number: if mod(Ysize,2)~=0 % modulus after division Ysize = Ysize - 1; Xsize = Xsize - 1; end % To produce image data in final output form: image_sequence = eval(file_info.name); % evaluate the string containing the matrix name to get the matrix data into a variable. %vectorised footage = mat2gray(double(cat(3,image_sequence(:,:,:)))); % Loop through frames: % for p = 1:numFrames % frame = image_sequence(:,:,p); % % frame = single(frame); % to class single. % frame = double(frame); % to class double. % frame = mat2gray(frame); % footage(p).image = frame; % % image data needs to be of class double for future operations (for findSpotCentre1frame.m). % end fData(1).numFrames = numFrames; fData(1).Xsize = Xsize; fData(1).Ysize = Ysize; end fData(1).numSeries = numSeries;
mat2gray each frame individually
for p = 1: fData(1).numFrames; footage(:,:,p) = mat2gray(footage(:,:,p)); end
Output info to command window:
disp(file) % write footage name to command window. fprintf('There are a total of %i series in this data file \nThe total number of frames in this image sequence is: %i \n ',fData(1).numSeries, fData(1).numFrames) %output total number of frames to command window.